Manufacturing data analysis device, system, and method

1 . A manufacturing data analysis device comprising:

a display device;

processing circuitry operably coupled to a manufacturing data database storing manufacturing data on a plurality of products; and

a storage storing a machine learning model trained to output a value of a degree of influence based on input of the manufacturing data, the output of the value of the degree of influence comprising calculating a bias in anomaly rates of the products and selecting a bias rate whose value is largest among calculated bias anomaly rates as the degree of influence,

wherein the processing circuitry is configured to:

acquire, from the manufacturing data database under a first acquisition condition, first manufacturing data including a manufacturing condition data group and a quality data group, the manufacturing condition data group including one or more pieces of manufacturing condition data on a value indicating a manufacturing condition for each of the products, and the quality data group including one or more pieces of quality data on a value indicating quality for each of the products;

calculate one or more first degrees of influence exerted by first manufacturing condition data included in the manufacturing condition data group on the respective pieces of quality data included in the quality data group, by performing analysis processing using the stored machine learning model with the first manufacturing data as input;

in a first case where one or more first degrees of influence out of the calculated one or more first degrees of influence satisfy a first determination condition:

generate first output data comprising image data displayable on the display device and including a content related to at least one of the first manufacturing condition data, one or more pieces of first quality data on which the first manufacturing condition data has exerted the first degrees of influence satisfying the first determination condition, or the first degrees of influence satisfying the first determination condition; and

calculate one or more second degrees of influence exerted by each piece of second manufacturing condition data on each of the pieces of first quality data, the second manufacturing condition data being included in the manufacturing condition data group and differing from the first manufacturing condition data; and

in a second case where one or more second degrees of influence out of the calculated one or more second degrees of influence satisfy a second determination condition, generate second output data comprising image data displayable on the display device and including a content related to at least one of one or more pieces of the second manufacturing condition data that have exerted the second degrees of influence satisfying the second determination condition, one or more pieces of the first quality data on which the one or more pieces of second manufacturing condition data have exerted the second degrees of influence satisfying the second determination condition, or the second degrees of influence satisfying the second determination condition,

wherein:

the image data comprising at least one of the first output data or the second output data includes a diagram in which a first axis indicates the manufacturing condition data and a second axis indicates the quality data,

the processing circuitry is configured to output the first output data and the second output data to the display device to control the display device to display the first output data and the second output data simultaneously, and

the processing circuitry is configured to change a display mode of the image data representing the first output data being displayed on the display device based on a value of the one or more first degrees of influence, and to change a display mode of the image data representing the second output data being displayed on the display device based on a value of the one or more second degrees of influence.

2 . The manufacturing data analysis device according to claim 1 , wherein the first acquisition condition is information for identifying each individual or group of individuals of the products.

3 . The manufacturing data analysis device according to claim 1 , wherein the first acquisition condition is at least one value included in at least one of the manufacturing condition data or the quality data.

4 . The manufacturing data analysis device according to claim 1 , wherein the first output data includes at least one of a name, a value, a representative value, a statistical value, a histogram, or a scatter diagram related to the first manufacturing condition data and the first quality data.

5 . The manufacturing data analysis device according to claim 1 , wherein the second output data includes at least one of a name, a value, a representative value, a statistical value, a histogram, or a scatter diagram related to the second manufacturing condition data and the first quality data.

6 . The manufacturing data analysis device according to claim 1 , wherein the processing circuitry is configured to determine that the one or more first degrees of influence satisfy the first determination condition in response to a determination that the one or more first degrees of influence out of the calculated one or more first degrees of influence are equal to or greater than a first threshold value.

7 . The manufacturing data analysis device according to claim 1 , wherein the processing circuitry is configured to determine that the one or more second degrees of influence satisfy the second determination condition in response to a determination that the one or more second degrees of influence out of the calculated one or more second degrees of influence are equal to or greater than a second threshold value.

8 . The manufacturing data analysis device according to claim 7 , wherein:

the processing circuitry is configured to determine that the one or more first degrees of influence satisfy the first determination condition in response to a determination that the one or more first degrees of influence out of the calculated one or more first degrees of influence are equal to or greater than a first threshold value, and

the second threshold value is equal to or greater than the first threshold value.

9 . The manufacturing data analysis device according to claim 1 , wherein the processing circuitry is configured to:

generate image data of a setting screen which enables a user to set parameters including at least one of the first acquisition condition, the first manufacturing condition data, the manufacturing condition data group, or the quality data group; and

output the image data of the setting screen to the display device to control the display device to display the setting screen.

10 . The manufacturing data analysis device according to claim 9 , wherein the processing circuitry is configured to generate a list of the parameters based on the manufacturing data, and to generate the image data of the setting screen to include the generated list.

11 . The manufacturing data analysis device according to claim 1 , wherein:

the processing circuitry is configured to:

determine a second acquisition condition, which is different from the first acquisition condition, based on the first manufacturing data; and

acquire the second manufacturing data including the manufacturing condition data group and the quality data group from the manufacturing data under the second acquisition condition; and

calculate the first degree of influence and the second degree of influence exerted by the second manufacturing condition data on each of the pieces of quality data, by performing the analysis processing using the stored machine learning model with the first manufacturing data and the second manufacturing data as input.

12 . The manufacturing data analysis device according to claim 1 , wherein the diagram is a histogram or a scatter diagram.

13 . The manufacturing data analysis device according to claim 1 , wherein the processing circuitry is configured to determine that the one or more first degrees of influence satisfy the first determination condition in response to a determination that the one or more first degrees of influence out of the calculated one or more first degrees of influence are equal to or greater than a first threshold value.

14 . The manufacturing data analysis device according to claim 13 , wherein generating the first output data comprises generating, as the image data, image data in which the quality data is highlighted, in response to a determination that a plurality of values related to the quality data are equal to or greater than a threshold value.

15 . The manufacturing data analysis device according to claim 1 , wherein the processing circuitry is configured to determine that the one or more second degrees of influence satisfy the second determination condition in response to a determination that the one or more second degrees of influence out of the calculated one or more second degrees of influence are equal to or greater than a second threshold value.

16 . The manufacturing data analysis device according to claim 15 , wherein generating the second output data comprises generating, as the image data, image data in which the quality data is highlighted, in response to a determination that a plurality of values related to the quality data are equal to or greater than a threshold value.

17 . The manufacturing data analysis device according to claim 1 , wherein the first output data includes, among the image data thereof, image data indicating a warning based on the one or more first degrees of influence.

18 . The manufacturing data analysis device according to claim 17 , wherein the first output data includes, among the image data thereof, a plurality of characters representing the warning or a highlighted frame border including the warning.

19 . The manufacturing data analysis device according to claim 18 , wherein the changing the display mode of the image data representing the first output data comprises:

modifying at least one of a font, a color, or a size of the characters being displayed, or

changing at least one of a type, a color, or a thickness of the frame border being displayed.

20 . The manufacturing data analysis device according to claim 1 , wherein the second output data includes, among the image data thereof, image data indicating a warning based on the second degrees of influence.

21 . The manufacturing data analysis device according to claim 20 , wherein the second output data includes, among the image data thereof, a plurality of characters representing the warning or a highlighted frame border including the warning.

22 . The manufacturing data analysis device according to claim 21 , wherein the changing the display mode of the image data representing the second output data includes:

modifying at least one of a font, a color, or a size of the characters being displayed, or

changing at least one of a type, a color, or a thickness of the frame border being displayed.

23 . The manufacturing data analysis device according to claim 1 , wherein:

the processing circuitry is configured to:

determine a second acquisition condition, which is different from the first acquisition condition, based on the first manufacturing data;

acquire the second manufacturing data including the manufacturing condition data group and the quality data group from the manufacturing data under the second acquisition condition; and

calculate another second degree of influence exerted by the second manufacturing condition data on each of the pieces of quality data, by performing the analysis processing using the stored machine learning model with the second manufacturing data as input.

24 . The manufacturing data analysis device according to claim 1 , wherein generating at least one of the first output data or the second output data comprises generating, as the image data, image data indicating a degree of reliability of each analysis result.

25 . The manufacturing data analysis device according to claim 1 , wherein:

the processing circuitry is configured to:

determine a second acquisition condition, which is different from the first acquisition condition, based on the first manufacturing data;

acquire the second manufacturing data including the manufacturing condition data group and the quality data group from the manufacturing data under the second acquisition condition; and

calculate another first degree of influence exerted by the first manufacturing condition data on each of the pieces of quality data, by performing the analysis processing using the stored machine learning model with the second manufacturing data as input.

26 . A manufacturing data analysis system comprising:

a manufacturing data database storing manufacturing data on a plurality of products;

a manufacturing data analysis device operably coupled to the manufacturing data database; and

a display device operably coupled to the manufacturing data analysis device,

wherein:

the manufacturing data analysis device includes (i) processing circuitry which is operably coupled to the manufacturing data database, and (ii) a storage storing a machine learning model trained to output a value of a degree of influence based on input of the manufacturing data, the output of the value of the degree of influence comprising calculating a bias in anomaly rates of the products and selecting a bias rate whose value is largest among calculated bias anomaly rates as the degree of influence,

the processing circuitry is configured to:

acquire, from the manufacturing data database under a first acquisition condition, first manufacturing data including a manufacturing condition data group and a quality data group, the manufacturing condition data group including one or more pieces of manufacturing condition data on a value indicating a manufacturing condition for each of the products, the quality data group including one or more pieces of quality data on a value indicating quality for each of the products;

calculate one or more first degrees of influence exerted by first manufacturing condition data included in the manufacturing condition data group on the respective pieces of quality data included in the quality data group, by performing analysis processing using the stored machine learning model with the first manufacturing data as input;

in a first case where one or more first degrees of influence out of the calculated one or more first degrees of influence satisfy a first determination condition:

generate first output data comprising image data displayable on the display device and including a content related to at least one of the first manufacturing condition data, one or more pieces of first quality data on which the first manufacturing condition data has exerted the first degrees of influence satisfying the first determination condition, or the first degrees of influence satisfying the first determination condition; and

calculate one or more second degrees of influence exerted by each piece of second manufacturing condition data on each of the pieces of first quality data, the second manufacturing condition data being included in the manufacturing condition data group and differing from the first manufacturing condition data; and

in a second case where one or more second degrees of influence out of the calculated one or more second degrees of influence satisfy a second determination condition, generate second output data comprising image data displayable on the display device and including a content related to at least one of one or more pieces of the second manufacturing condition data that have exerted the second degrees of influence satisfying the second determination condition, one or more pieces of the first quality data on which the one or more pieces of second manufacturing condition data have exerted the second degrees of influence satisfying the second determination condition, or the second degrees of influence satisfying the second determination condition,

the image data comprising at least one of the first output data or the second output data includes a diagram in which a first axis indicates the manufacturing condition data and a second axis indicates the quality data,

the processing circuitry is configured to output the first output data and the second output data to the display device to control the display device to display the first output data and the second output data simultaneously, and

the processing circuitry is configured to change a display mode of the image data representing the first output data being displayed on the display device based on a value of the one or more first degrees of influence, and to change a display mode of the image data representing the second output data being displayed on the display device based on a value of the one or more second degrees of influence.

27 . A manufacturing data analysis method executed under control of processing circuitry operably coupled to a display device, a storage, and a manufacturing information database storing manufacturing data on a plurality of products, the method comprising:

storing, in the storage, a machine learning model trained to output a value of a degree of influence based on input of the manufacturing data, the output of the value of the degree of influence comprising calculating a bias in anomaly rates of the products and selecting a bias rate whose value is largest among calculated bias anomaly rates as the degree of influence;

acquiring, from the manufacturing information database under a first acquisition condition, first manufacturing data including a manufacturing condition data group and a quality data group, the manufacturing condition data group including one or more pieces of manufacturing condition data on a value indicating a manufacturing condition for each of the products, the quality data group including one or more pieces of quality data on a value indicating quality for each of the products;

calculating one or more first degrees of influence exerted by first manufacturing condition data included in the manufacturing condition data group on the respective pieces of quality data included in the quality data group, by performing analysis processing using the stored machine learning model with the first manufacturing data as input;

in a first case where one or more first degrees of influence out of the calculated one or more first degrees of influence satisfy a first determination condition:

generating first output data comprising image data displayable on the display device including a content related to at least one of the first manufacturing condition data, one or more pieces of first quality data on which the first manufacturing condition data has exerted the first degrees of influence satisfying the first determination condition, or the first degrees of influence satisfying the first determination condition; and

calculating one or more second degrees of influence exerted by each piece of second manufacturing condition data on each of the pieces of first quality data, the second manufacturing condition data being included in the manufacturing condition data group and differing from the first manufacturing condition data; and

in a second case where one or more second degrees of influence out of the calculated one or more second degrees of influence satisfy a second determination condition, generating second output data comprising image data displayable on the display device and including a content related to at least one of one or more pieces of the second manufacturing condition data that have exerted the second degrees of influence satisfying the second determination condition, one or more pieces of the first quality data on which the one or more pieces of second manufacturing condition data have exerted the second degrees of influence satisfying the second determination condition, or the second degrees of influence satisfying the second determination condition,

wherein:

the image data comprising at least one of the first output data or the second output data includes a diagram in which a first axis indicates the manufacturing condition data and a second axis indicates the quality data, and

the method further comprises:

outputting the first output data and the second output data to the display device to control the display device to display the first output data and the second output data simultaneously, and

changing a display mode of the image data representing the first output data being displayed on the display device based on a value of the one or more first degrees of influence, and changing a display mode of the image data representing the second output data being displayed on the display device based on a value of the one or more second degrees of influence.